Phase 3 Failure for Knee Osteoarthritis Cell and Gene Therapy as TG-C Fails to Outperform Saline

A single injection of a cell and gene therapy that could provide long-term pain relief and slow knee joint degeneration could have filled the treatment gap between medication and joint replacement surgery. However, Kolon TissueGene’s TG-C failed to demonstrate superiority to saline in the U.S. phase 3 ACTiVION-II trial, marking a critical setback for a development path pursued over many years.

This randomized, double-blind, placebo-controlled study enrolled 531 patients with Kellgren–Lawrence grade 2 or 3 knee osteoarthritis at 27 centers in the United States and compared a single ultrasound-guided intra-articular injection of TG-C with saline. At month 12 after treatment, neither of the two co-primary endpoints—the visual analog scale for pain (VAS) and the WOMAC total score, which comprehensively assesses pain, stiffness, and ability to perform daily activities—showed a statistically significant difference. All prespecified key secondary endpoints were also missed.

The figures show that symptoms improved markedly in both groups, but there was almost no difference between them. The mean changes in VAS scores in the TG-C and placebo groups were −38.7 and −39.2 points, respectively, a difference of only 0.5 points, with a p-value of 0.8322. The mean changes in WOMAC total scores were −27.61 and −26.54 points, respectively, a difference of −1.07 points, with a p-value of 0.5701. The company believes a greater-than-expected placebo response may have affected the results, but this explanation still requires validation through a full analysis of the data and cannot replace the conclusion that the trial failed to meet its endpoints.

TG-C combines donor-derived cells with genetically engineered cells capable of expressing TGF-β1 and is designed to act locally within the joint after a single injection. The company hopes it can not only relieve symptoms but also affect disease progression. However, the results announced this time are topline results and are not sufficient to determine whether specific patient subgroups may benefit, nor do they provide sufficient evidence to support structural improvement of the joint.

Regarding safety, no new or unexpected signals have been observed to date. The incidence of treatment-emergent adverse events was 83.9% in the TG-C group and 78.1% in the placebo group, with most classified as grade 1 or grade 2. The proportions of serious adverse events were 10.3% and 8.6%, respectively. However, the absence of new warning signals does not mean that long-term risks have been ruled out. Trial registration data show that, in addition to 24 months of efficacy and safety follow-up, annual questionnaires are planned after dosing to monitor for cancer for up to 15 years.

TG-C’s fate has not yet been fully determined. Kolon TissueGene expects to announce results from another independent phase 3 trial, ACTiVION-I, in October 2026, and will decide the program’s next steps afterward. If the second study also fails to show a difference from placebo, TG-C’s U.S. development and commercialization strategy will inevitably require substantial reassessment. If the results differ, the company will need to explain why the two large trials diverged and demonstrate that the difference was not due to chance or the method of analysis.